116s
PHILIPL. SOUTHWICK AND
removed b y filtration. The filtrate was concentrated in vacuo t o about 3 ml., cooled in a n ice-bath, and filtered. The filtrate was evaporated to dryness in vacuo. To the oily residue was added 20 ml. of acetone, followed by 60 nil. of ether, and the solution was stored S hours at 0". The product was collected by filtration. There was obtained 1.36 g. of colorless needles, which, upon two crystallizations from methanol: acetone: isopropyl ether, yielded 1.06 g. of analytically pure material. Four of the eight isomers (LDL,DLD,LDD, D L L ) failed to crystallize from acetone or from acetone: ether solution; instead semi-solid products were obtained (Table 111). Valylvalylvaline Tripeptide Isomers. D-Valyl-D-valyl-Dvaline.-O.500 g. (1.36 mmoles) of D-valyl-D-valpl-D-valine methyl ester hydrochloride was suspended in 11 ml. of 0.5 N potassium hydroxide, stored 75 minutes at 3 7 O , chilled, and washed with three 2-ml. portions of ethyl acetate. The p H of the clear aqueous solution was adjusted to 5.0
JOSEPH
CASANOVA, JR.
1701.
80
with dilute acetic acid and the resulting solution was concentrated in vacuo to about 2 ml., 25 ml. of ethanol was added and the crystallizing solution was stored overnight at 0". The crystalline product was collected by filtration, and recrystallized from 5 ml. of water and 30 ml. of ethanol; yield) of fine colorless plates was isolated. 0.371 g. ('is$& Flame photometry revealed less than 0.1% potassiurri. The product was dried to constant weight at 1 mm. a t loo", over P206,then exposed to room humidity until constant weight was reached. The tripeptides were hygroscopic and moistures were determined each time samples were weighed for characterization. The tripeptide isomers L L L , DDL and LLD were obtained in the same manner. The four tripeptide methyl ester hydrochlorides which failed to crystallize were saponified as above and yielded crystalline peptides (Table IV). PASADENA 4, CALIF.
[ CONTRIIlUrION FROM THE DEPARTMENT O F CHEMISTRY, CARNEGIE INSTITUTE OF
TECHNOI.OGY]
A New Synthesis of dl-Vasicine and a Methoxy Analog' BY PHILIPI.. SOUTHWICK AND JOSEPH CASANOVA, JR.* RECEIVED SEPTEMBER 18, 1957 dl-Vasicine has been prepared from o-nitrotoluene by a reaction sequence in which o-nitrobenzylamine hydrochloride ethyl 8-(o-nitrobenzy1amino)-propionatehydrochloride, l-(o-nitrobenzyl)-4-carbethoxy-2,3-dioxopyrrolidine,I-(o-nitrobenzyl)-2,3-dioxopyrrolidineand 1-(o-nitrobenzyl)-3-hydroxy-2-oxopyrrolidinewere the intermediates isolated and purified. A strictly parallel series of reactions, with only very slight modification of the details of procedure, served for the preparation ("dZ-6-methoxyvasicine"), from 3-methyl-4-nitroanisole. of a new vasicine analog, dZ-3-hydroxy-G-methoxypeg-9-ene
Work on the chemistry of vasicine (Ia) by several There have been indications in the literatures of groups of workersa culminated in the synthesis of interesting physiological effects, such as bronchodithe alkaloid by Spath, Kuffner and Platzer4 in lator activity, produced by vasicine, and for this 1935. The first synthesis depended on the prep- reason we have been prompted to undertake the aration of methyl a-hydroxy-y-aminobutyrate, synthesis of closely related analogs of vasicine in which was treated with o-nitrobenzyl chloride to the hope of obtaining compounds of increased acproduce a low yield (ca. 7%) of l-(o-nitrobenzyl)-3- tivity. In preparation for this effort we have inhydroxy-2-oxopyrrolidine (IIa). Compound IIa, vestigated a new route to vasicine which proceeds when treated with stannous chloride in hydro- through the final intermediate I I a of the first synchloric acid, underwent reduction of the nitro thesis by Spath, Kufker and P l a t ~ e rbut , ~ utilizes a group and spontaneous cyclization with dehydra- different synthesis for this key compound. It was tion to give a 35% yield of dl-vasicine. Later desired to achieve a preparative route which might (1936) Spath and Platzefl described a more conven- be expected to be reliable when used on a relatively ient procedure whereby a-hydroxybutyrolactone large scale and applicable without extensive modifi(111) was prepared from butyrolactone and heated cation when applied to the preparation of vasicine to a high temperature with o-aminobenzylamine analogs It was also important that the synthetic (IV) to give dl-vasicine, with a yield in the final sequence serve as an adequate structure proof for the products. These requirements led us to seek to step of 20%. avoid those cyclization methods used in the earlier syntheses which had produced low yields, necessitated the use of complex procedures for the isolation of the product, or utilized pairs of reactants whose manner of reacting might be considered unIV I11 OH certain. The reaction sequence adopted is shown in Chart (1) This investigation was supported in part by a research graut I. If vasicine analogs substituted in the aromatic (RG-4371) from the Division of Research Grants, National Institutes o f Health, Public Health Service. ring were to be obtained by this route, a reasonably (2) Institute Fellow in Organic Chemistry, 1956-1957. This paper convenient method for the preparation of ring-subis abstracted from a thesis submitted by Joseph Casanova, Jr., in stituted o-nitrobenzylamines (V) was needed. partial fulfillment of the requirements for the degree of Doctor of Chart I1 outlines a sequence which may prove to Philosophy at the Carnegie Institute of Technology, May, 1957. (3) Reviews of this work include those (a) by H. T. Oppenshaw be applicable to the preparation of a number of such (R. H. F. Manske and H. L. Holmes, "The Alkaloids," Vol. 111, compounds. It was tested initially in the preparaAcademic Press, Inc., New York, N. Y . , 1953, pp. 101-111); (b) tion of o-nitrobenzylamine itself (Va). o-Nitroby T. A. Henry, "The Plant Alkaloids," 3rd ed., P . Blakiston's Sons and Co., Inc., Philadelphia, Pa., 1939, pp. 544-548; and (c) by E. toluene was brominated with N-bromosuccinimide SpZth, Monalsh., 74, 115 (1938). (4) E. Spilth, F. Kuffner and H. Platrer, Bcr., 68, 699 (1935). (5) E.Spilth and N. Platzcr. ibid., 69, 266 (1988).
(6) (a) D. Hooper, Phorm. J . , [3] 18, 841 (1888); (b) R.M.Chopn and S. Ghosh, Indian Mld. Gas., 60,864 (1926): C. A , , 19,8323 (1926).
March 5, 1055
dl-VASICINE AND
A METHOXY
ANALOG
1169
CHARTI
VIIa,b
VIa,b
1Xa.b
VIIIa,b
I
Xa,b
1Ia.b
as described by Kornblum and Ifflz~nd,~ and the resulting crude o-nitrobenzyl bromide (XIIa) was treated with hexamethylenetetramine to produce o-nitrobenzylhexaminium bromide (XIIIa).8 This salt was then converted first into o-nitrobenzylaminomethylol sulfite (XIVa)g by treatment with water and sulfur dioxide, and thence into onitrobenzylamine hydrochloride (XVa) by hydrolysis of the methylol sulfite (XIVa) with hydrochloric acid. The conversion of X I a into XVa was only 19%, but the procedure proved to be reliable and suitable for use on a reasonably larger scale. The intermediate conversion of the hexaminium salt X I I I a to the methylol sulfite XIVa, as described by Reichert and D ~ r n i s provided ,~ an effective means of avoiding the loss of product through the formation of o-nitrobenzaldehyde. CHART I1
SIa, R = H b, R OCHI
SIIa,b,
H20 '
SIIIa,b
SIVa,b
XVa,b
The conversion of o-nitrobenzylamine (Va) to the hydroxypyrrolidone I I a was accomplished largely by methods developed in the course of earlier investigations of the chemistry of pyrrolidine derivatives.l0 Thus, the first step (reaction 1, Chart I) (7) N. Kornblum and D. C.Iffland,THIS J O U R N A L . 71, 2137 (1949). (8) S. J. Augyal, P. J. Morris, J. P. Tetne and J. C. Wilson, J . Chcm. SOC.,2141 (1950). (9) B. Reichert and W. Dornis, Arch. Pharm., 282, 100 (1944); C. A . , 46, 1969f (1951). (IO) (a) P. L. Southwick and L. L. Seivard, THISJOURNAL, 11, 2532 (1949); (b) P. L. Southwick and R . T. Crouch, ibid., 76, 3413
Ia,b
OH
was addition of o-nitrobenzylamine to ethyl acrylate to produce ethyl 0-(o-nitrobenzy1aminc)-propionate (VIa) in a yield of 81%. This compound was purified and characterized in the form
